Engineering Escherichia coli K-­12 for the secretion of single domain antibodies against attaching and effacing bacterial pathogens and for the injection of proteins of therapeutic potential into human cells

Abstract

The attaching and effacing (A/E) bacterial pathogens infect the gastrointestinal tract of humans and other mammals after ingestion of contaminated food or water and cause persistent diarrhoea and other important diseases (e.g. haemolytic uremic syndrome, HUS) worldwide. Prototypical A/E pathogens are the enteropathogenic Escherichia coli (EPEC) and enterohaemorrhagic E. coli (EHEC) strains, which infect humans, as well as the mouse-­‐restricted pathogen Citrobacter rodentium (CR). These pathogens contain a common type III secretion system (T3SS): a macromolecular protein complex (the injectisome) assembled in the bacterial cell envelope that protrudes to the extracellular milieu with a filament of polymerized EspA. The T3SS allows translocation (injection) of a repertoire of bacterial proteins (called effectors) into the cytoplasm of the host enterocytes through the translocon subunits EspB and EspD, which insert in the host cell membrane. The effectors subvert multiple cellular functions and cause the effacement of the intestinal microvilli (A/E lesions) and the disruption of the intestinal epithelial barrier. Among them, the translocated intimin receptor (Tir) inserts in the host cell plasma membrane and is recognised by Intimin (Int), an outer membrane adhesin exposed on the bacterial cell surface. The Int:Tir interaction promotes the intimate attachment of the bacterium to the enterocyte and the polymerization of actin filament bundles (called “pedestals”) underneath the attached bacterium. Effective treatments to combat A/E pathogens are needed, since antibiotics activate the expression of the life-­‐threatening Shiga-­‐like toxins (Stx) from integrated pro-­‐phages, which are present in the more virulent strains -­‐ such as EHEC O157:H7. In this work we have assessed whether single domain antibodies (sdAbs) from camelids (also known as nanobodies or VHHs) binding essential proteins for A/E lesion formation (i.e., EspA, Int, Tir) could act as potential therapeutic agents against EHEC O157:H7 infections if secreted from non-­‐pathogenic E. coli strains. In addition, we have explored the biotechnological use of the filamentous T3SS of the A/E pathogens to inject VHHs and other proteins with therapeutic potential into the cytoplasm of human cells using a non-­‐ pathogenic E. coli K-­‐12 strain, engineered to express functional EPEC injectisomes. We have immunized a dromedary with purified EspA, Int280 and TirM (the protein domains involved in Int:Tir interaction) of EHEC to create a library of VHH genes. Selected VHHs clones from this library were secreted by commensal E. coli K-­‐12 strains carrying the α-­‐haemolysin (HlyA) secretion system and purified from the extracellular medium to characterise their binding and inhibitory properties of Int:Tir interaction in vitro. A high-­‐affinity VHH clone recognising TirM -­‐ named TD4 -­‐ and blocking Int:Tir interaction was found to interfere with the formation of actin pedestals on HeLa cells infected with EHEC. We have demonstrated that TD4 effectively competes with Int280 for the binding of Tir, since it has higher affinity towards TirM and recognises an epitope that overlaps with the necessary residues for the Int:Tir interaction. This VHH showed high specificity towards TirM of EHEC, not binding TirM of EPEC and only weakly with TirM of C. rodentium. With the aim to establish an in vivo mouse model for the evaluation of the A/E inhibition by TD4, we generated a C. rodentium strain expressing EHEC proteins Int, Tir, the multicargo chaperone CesT and the Tir-­‐coupling protein effector TccP. The resulting strain (CR-­‐EHEC) was capable of forming actin pedestals on HeLa cells, colonizing the mouse intestinal tract and inducing crypt hyperplasia, similarly to wild type C. rodentium. Importantly, we confirmed that TD4 also inhibits the formation of actin pedestals on HeLa cells induced by CR-­‐EHEC. These results leave open the possibility of testing the activity of TD4 against the formation of A/E lesions in vivo, which could represent the basis of a future EHEC infection treatment to control the outbreaks of this pathogen. In addition, we have addressed the biotechnological use of the filamentous T3SS of A/E pathogens for the injection of sdAbs and other heterologous proteins of therapeutic potential into the cytoplasm of human cells. The genes encoding the T3SS of EPEC are found in a 35 kb chromosomal island, called the Locus of Enterocyte Effacement (LEE). LEE is organized in various large operons (LEE1 to LEE5) and shorter transcriptional units (e.g. escD) and also contains genes encoding some effectors, their chaperones, transcriptional regulators, a muramidase and intimin. We aimed to engineer the expression of all known necessary genes for the assembly of a functional T3SS (27 genes in EPEC) in a non-­‐pathogenic E. coli strain (e.g. K-­‐12) under the control of the inducible promoter Ptac. We organised these genes in five engineered transcriptional units -­‐ operons eLEE1 to eLEE4 and eEscD -­‐ lacking effectors and transcriptional regulators. These engineered operons were sequentially integrated in specific sites of the genome of E. coli K-­‐12 corresponding with fimbrial and afimbrial adhesins. The resulting non-­‐pathogenic strain, referred to as Synthetic Injector E. coli (SIEC), was demonstrated to express the T3SS genes upon IPTG addition, to assemble functional injectisomes on its envelope and to be able to secrete the translocator proteins EspA, EspB, EspD. Interestingly, we found that the expression of the T3SS genes partially interfered with the assembly of the flagellum in E. coli K-­‐12, which can be explained considering the high identity between components of the T3SS and the flagellum. Lastly, we showed that a SIEC strain carrying an additional engineered operon encoding Int, Tir and CesT (eLEE5), was able to inject Tir to the cytoplasm of HeLa cells and to reproduce the intimate attachment of the bacterium to the host cell and the formation of actin pedestals. These results open the possible biomedical application of SIEC-­‐derived strains to inject proteins of therapeutic potential (e.g. VHHs, enzymes, transcription factors, toxins, effectors) against diverse human diseases (e.g. cancer, chronic inflammation).